cancel
Showing results for 
Search instead for 
Did you mean: 

ADSO Configuration - Master Data Check

Former Member
0 Kudos

Hi,

In ADSO we can select configuration for each characteristic to specify Master Data Check - see below image.

I found that SID's are generated for that characteristic only if option 3 and 4 (i.e. Check during load/activation or Persist SID in DataSource) are selected. But, if I select option 2 i.e. Master Data Check During Reporting, I am unable to select the values loaded in selection screen of report. Also, I cannot do a filter on those values in output as it gives error master data does not exist.

I am wondering why this option is given and why is it at characteristic level and not ADSO level. I have 30 chars in my ADSO and have to select it individually. In what scenario will I not want to create SID? Or are SID's optional now on BW on HANA?

Thanks,

Anoop

Accepted Solutions (1)

Accepted Solutions (1)

former_member216578
Participant
0 Kudos

Hi, Anoop.

Since HANA has a column store it uses a dictionary compression for columns of column tables.

Thus a column table is internally represented as a star: all distinct values in every column are sorted and saved in (key, value) pair in column dictionary (and uses sufficient and necessary key length to encode all values with the lowest overhead), and column vector keeps that keys instead of real values. So internally your table is already have a star schema representation and you don't need to create SIDs. I've already answered a similar question.

Regarding consistency check. You can load your master data independently and I think you should be able to select loaded MD values for filters. So if it's not present you'll get an error. MD check during Load/Activation increases a load time. I think it's set at characteristic level because your characteristics are of different cardinality, so you can decide to change some of them.

Regards,

Andrey.

Former Member
0 Kudos

Hi Andrey,

Thanks for your detailed explanation.

I need Billing document number and contract number in selection screen. Both these characteristics do not have any master data enabled so I cannot load master data separately into them to generate SID. So 'Check Master Data during data load' in ADSO for those characteristic has to be enabled mandatorily in my case.

Essentially, every characteristic that has attribute and master data is loaded will have SID's generated (am I right?). Also, every characteristic that you need to use in selection screen needs to have SID. With this, I guess for most characteristics SID's will be needed.

Also, what if I do not generate SID for say 'Document Number' and after 1 year customer wants to add 'Document Number' in selection screen. This will demand complete reload of data to ADSO. So, is it not safer to always generate SID for all characteristics?

I understand that in BW on HANA SID's is not mandatory, but if above is true, then SID is still needed in most cases. Do share your view.

Thanks,

Anoop Sahu

former_member216578
Participant
0 Kudos

Hi, Anoop.

I've figured out how to maintain MD check property at DSO level. See image below.

Regarding filters. If infoobject represents a system-generated ID it's really meaningless. If you delete some document and create it for the second time it will get another ID, but business entity (document printed on the paper) will be the same. It's not a good practice to allow users to remember this IDs. If you have some business numbers like document regnumber, then of course you can have a requirement to use it for filter. And you have no choice - you should generate SIDs to have an object which holds available values.

Regards,

Andrey.

Former Member
0 Kudos

Thanks a lot Andrey.

Seems my Studio needs an update. I do not see 'Maintain Properties' button in ADSO. I am currently on Version: 2.1.6

former_member216578
Participant
0 Kudos

Hi, Anoop.

I'm using Eclipse Luna with BWMT 1.12.10. Try to update your BWMT or another Eclipse version.

Regards,

Andrey.

Answers (0)